In this paper, we investigate the communication characteristics of the Message Passing Interface (MPI) implementation of the NAS parallel benchmarks and study the effectiveness of compiled communication for MPI programs. Compiled communication is a technique that utilizes the compiler knowledge of both the application communication requirement and the underlying network architecture to significantly optimize the performance of communications whose information can be determined at compile time (static communications). The results indicate that compiled communication can be applied to a large portion of the communications in the benchmarks. In particular, the majority of collective communications are static
Clusters of workstations are a popular platform for high-performance computing. For many parallel ap...
We discuss the design and high-performance implementation of collective communications operations on...
The message passing interface standard released in April 1994 by the MPI Forum [2], defines a set of...
In this paper we describe a compiler framework which can identify communication patterns for MPI-bas...
In this paper we describe a compiler framework which can identify communication patterns for MPIbase...
Compiled communication has recently been proposed to improve communication performance for clusters ...
Compiled communication has recently been proposed to improve communication performance for clusters ...
Parallel computing on clusters of workstations and personal computers has very high potential, since...
In order for collective communication routines to achieve high performance on different platforms, t...
We have implemented eight of the MPI collective routines using MPI point-to-point communication rou...
The Message-Passing Interface (MPI) is a widely-used standard library for programming parallel appli...
Parallel computing on clusters of workstations and personal computers has very high potential, sinc...
Most parallel and sequential applications achieve a low percentage of the theoretical peak performan...
Collective communication is an important subset of Message Passing Interface. Improving the perform...
A benchmark test using the Message Passing Interface (MPI, an emerging standard for writing message ...
Clusters of workstations are a popular platform for high-performance computing. For many parallel ap...
We discuss the design and high-performance implementation of collective communications operations on...
The message passing interface standard released in April 1994 by the MPI Forum [2], defines a set of...
In this paper we describe a compiler framework which can identify communication patterns for MPI-bas...
In this paper we describe a compiler framework which can identify communication patterns for MPIbase...
Compiled communication has recently been proposed to improve communication performance for clusters ...
Compiled communication has recently been proposed to improve communication performance for clusters ...
Parallel computing on clusters of workstations and personal computers has very high potential, since...
In order for collective communication routines to achieve high performance on different platforms, t...
We have implemented eight of the MPI collective routines using MPI point-to-point communication rou...
The Message-Passing Interface (MPI) is a widely-used standard library for programming parallel appli...
Parallel computing on clusters of workstations and personal computers has very high potential, sinc...
Most parallel and sequential applications achieve a low percentage of the theoretical peak performan...
Collective communication is an important subset of Message Passing Interface. Improving the perform...
A benchmark test using the Message Passing Interface (MPI, an emerging standard for writing message ...
Clusters of workstations are a popular platform for high-performance computing. For many parallel ap...
We discuss the design and high-performance implementation of collective communications operations on...
The message passing interface standard released in April 1994 by the MPI Forum [2], defines a set of...